Protein sphericity

Protein sphericity

Mathematically representing and classifying protein 3D structures in biology is a challenging problem. The structure determines the function of proteins. Therefore, defining the global and local descriptors of the spatial architecture of proteins is a primary technology.

All the biological organisms on Earth consist of molecules. Among others, protein is the most important unit of life. Proteins consist of amino acids
and amino acids consist of atoms. Proteins are a versatile and diverse structures; small proteins contain several hundreds of atoms while large proteins contain millions of atoms. Not only that, proteins have an extremely diverse repetoire of 3D structures. For example, some proteins are extremely long fibers while some others are almost perfect spheres. Depending the overall shapes, the proteins function completely differently. Therefore, representing protein shapes is an important primary research problem in biology. Classifying proteins or mapping the protein universe has been a core research area in structural biology in the last four decades. Proteins are usually classified by their secondary structure contents such as alpha-helix, beta-sheets, and coils. This secondary structure based classification is useful for studying the evolution of proteins. For classifying proteins and mapping their structure-function relationships, researchers use structure representation descriptors. As proteins are dynamic and versatile molecules, there can be different ways of representing the basic 3D architecture. Until now, there are many physical representation models for protein structures. One of the most popular and convenient models is a hard-sphere model with a fixed
radius, the van der Waals radius, for each type of atom. Representing such physical properties is the first step of protein structure analysis and the representation should be easily defined and computed. 

See also
Sphericity